The first two forms of information warfare discussed here deal with attacks either on systems (C2 warfare) or by systems (IBW). The third form is EW, or operational techniques: radioelectronic and cryptographic, thus war in the realm of communications. EW attempts to degrade the physical basis for transferring information, while cryptographic warfare works between bits and bytes.

Neither type of EW is truly new. In tandem, they underlay Britain's success in defending its island against the Luftwaffe. In recent years, as information warfare has acquired a certain cachet, efforts have been made to reinvent EW under this new moniker. Note 24 Its supposed current rise in status is occurring just as technologies are being developed that will favor the bits (like the bomber of yore) getting through.

Antiradar

Note 25 A large portion of the EW community deals with radars (both search and target) and worries about jamming and counterjamming. Offense and defense keep coming up with new techniques. Traditional radars generate a signal at one frequency; knowing the frequency makes it easy to jam a return signal. More modern radars hop from one outgoing frequency band to the next. To counter radars, today's jammers must be able to acquire the incoming signal, determine its frequency, tune the outgoing jamming signal accordingly, and send a blur back quickly enough to minimize the length and strength of the reflected signal. Jamming aircraft that are riding in formation with attack aircraft often wipe out return signals (which weaken as the fourth power of the distance between radar and target) by overpowering them, but doing so makes jammers very visible so they must protect themselves. Coalition forces in the Gulf developed new synergies using jamming aircraft en masse. Radars make themselves targets because of their outgoing signals; antiradiation missiles (e.g., the HARM) force radars either to be turned off or to rely on chirping and sputtering. The aborted Tacit Rainbow missile was designed to loiter in an attack area until a radar turned itself on; the outgoing signal gave the missile an incoming beacon, and away it went. As digitization improves, radar can acquire a target by generating a transient pulse and analyzing the return signal before a false jamming signal overwhelms the reflection.

The cheaper digital manipulation becomes, the more logic favors the separation of an emitter from a collector. Emitters, the targets of antiradiation missiles, would proliferate, to ensure the survival of the system and to act as sponges for expensive missiles. The missiles would create a large virtual dish out of a collection of overlapping small ones. Because outgoing signals will be more complex, collection algorithms too will grow in complexity, but the ability of jammers to cover the more complex circle adequately may lag. Dispersing the collection surface will also make radars less inviting targets.

Anticommunications

EW against communicators is generally more difficult to wage than EW against radars. The signal strength of communications weakens with the distance to the transmitter squared (versus the fourth power with radar). While radars try to illuminate a target (and therefore send a beam into the assets of the other side), communicators try to avoid the other side entirely and thus point in specific directions. Communicators move toward frequency- hopping, spread-spectrum, and code-division multiple access (CDMA) technologies, which are difficult to jam and intercept. Communications to and from known locations (e.g., satellites, UAVs) can use digital technologies to focus on frontal signals and discard jamming that comes from the sides. Digital compression techniques coupled with signal redundancy mean that bit streams can be recovered intact, even if large parts are destroyed.

EW is also used to geolocate the emitter. The noisier the environment, the more difficult the task. One defense is to multiply sources of background electronic clutter shaped to foil intercept techniques that rely on distinguishing real signal patterns. Note 26 A thorough job, of course, requires expending resources to scatter emitters in areas where they may plausibly indicate military activity. Doing so diverts resources from other missions.

As suggested above, the work of finding targets is likely to shift from manned platforms to distributed systems of sensors. Despite the impending necessity of distributed systems, their Achilles' heel is the need for reliable, often heavily used communications links between many sensors, command systems, and dispersed weapons. Note 27 In sensor-rich environments, EW -- expressed by jamming or by soft- kill -- can assume a new importance. Interference with communications from local sensors, for instance, can create virtual blank areas through which opposing systems can move with less chance of detection. The success of this tactic critically depends on the architecture of the distributed sensor system to be disrupted. A system that relies exclusively on distributed local sensors (intercommunicating or relaying signals by low power to switches) is the most vulnerable. A system that interleaves local and stand-off sensors, particularly where coverage varies and overlap is common, is more robust.

Cryptography

By scrambling its own messages and unscrambling those of the other side, each side performs the quintessential act of information warfare, protecting its own view of reality while degrading that of the other side. Although cryptography continues to attract the best minds in mathematics, sadly for an otherwise long and glorious history, contests in this realm will soon be only of historical interest.

Decoding computer-generated messages is fast becoming impossible. The combination of technologies such as the triple- digital encryption standard (DES) for message communication using private keys, and public key encryption (PKE) for passing private keys using public keys (so set up communications remain in the clear) will probably overwhelm the best code-breaking computers. The basic mathematics is simple: for any key length x, for DES data encryption the power required to break the codes Note 28 is A*Nx (where x is the key length, A is positive, and N exceeds 1) and the power required to make the codes is B*Xm (where B is positive and M exceeds 1). Regardless of the quantity of A, B, M, and N, as soon as x exceeds some number, breaking a code is harder than creating one and becomes increasingly harder as x grows.

Although encryption is spreading on the Internet and all communications are going digital, the transition to ubiquitous encryption will take time. Analog will certainly persist in legacy systems, although its lifetime is limited. Cheap encryption, coupled with signal-hiding techniques such as spread-spectrum and frequency-hopping, will seal the codebreaker's fate.

Digital technologies will make spoofing -- substituting deceptive messages for valid ones -- nearly impossible. Digital- signature technologies permit recipients to know both who (or what) sent the message and whether the message was tampered with. Unless the spoofer can get inside the message-generation system or the recipient cannot access a list of universal digital keys (e.g., updates are unavailable to that location), the odds of a successful spoof are becoming quite low. Note 29